The Net Built to Survive the Bomb — ARPANET's First Cry, 1969

Today the internet is simply there, as ordinary as water or electricity. Yet it is surprisingly little known that its beginning lay at the crossing of two things: the terror of the Cold War, and the modest wish of a handful of researchers. A communications network that could survive a nuclear war. A way to share expensive machines standing far apart. As those two motives tangled together, a single wire that would change the world was drawn in 1969. The story begins with that first cry.

A Net to Survive a Nuclear Strike

It all began with a single fear. In the late 1950s, the United States Air Force needed a communications network that could survive a Soviet nuclear strike. The old telephone network ran on ‘circuit switching’ — for the length of a call, it seized an entire dedicated line joining two points. But this was a system weak to a single point of failure: destroy one point along the way, and the whole thing was severed. It could not withstand a nuclear war.

To this hard problem, Paul Baran of the RAND Corporation gave an answer. His 1962 concept ran like this. Do not gather the net at a centre; scatter it across countless paths. Then break the message into fine ‘message blocks’, and send each one independently, along whatever path lies open. Even if part is destroyed, the information reaches its destination by way of the surviving paths. In Britain, Donald Davies arrived independently at nearly the same idea in 1965, and gave it the name ‘packet switching’ — the very technology that still carries the world’s data communications today.

The Other Truth: It Was Not ‘For the Bomb’

Here a stubborn misunderstanding must be cleared away. The story that ARPANET itself was ‘built to survive a nuclear war’ is, in fact, not accurate.

It was Baran’s RAND research, not ARPANET, that tied packet switching to nuclear war. The motive that actually gave birth to ARPANET, at the United States’ Advanced Research Projects Agency (ARPA), was something plainer and more pressing. Charles Herzfeld, ARPA’s director at the time, testified to it plainly: ARPANET was not begun to build a command chain that could survive a nuclear strike. The true reason was a certain frustration — why could researchers far apart not use the handful of expensive research mainframes scattered across the country? A wish to share computing resources: that was the real point of departure.

The first to put that dream into words was J.C.R. Licklider, a psychologist and engineer. In 1963 he wrote to his ARPA colleagues of a grand vision — an ‘Intergalactic Computer Network’ in which distant computers would converse with one another. He himself left ARPA before the net stirred to life, but that vision was handed on to Bob Taylor, and on toward its realization in 1969.

‘LO’ — The Two Letters That Changed the World

In 1969, the vision finally became real. ARPA commissioned the firm BBN to build the network’s relay device, the Interface Message Processor (IMP) — the prototype of the later router — and the first four sites were chosen. UCLA (Leonard Kleinrock’s Network Measurement Center), the Stanford Research Institute, SRI (Douglas Engelbart), UC Santa Barbara, and the University of Utah.

On the night of October 29 that year, history moved. Charley Kline, a student programmer at UCLA, kept contact by telephone as he typed ‘LOGIN’ into SRI’s computer. ‘Can you see the L?’ ‘I see it.’ ‘And the O?’ ‘I see it.’ But the instant he typed the ‘G’, the system on SRI’s side crashed. And so the first message ever sent over ARPANET became, unintentionally, only the two letters ‘LO’ — as though calling out to the world, ‘Lo!’ About an hour later the system was restored, and the login succeeded. By December 5, the net of four sites was complete.

There was no grand fanfare. Yet this broken two-letter message was, beyond any doubt, the first voice of the vast network on which you are reading these very words today.

The net that linked these four sites would soon turn toward a greater dream. A net of radio, a net of satellites — to bind into one all the networks scattered across the world, each built to a different specification. But for that, it needed a common language that belonged to no one, one that could make itself understood to any counterpart at all. The story continues into the idea that truly created the internet.